How to generate stats

Activity and Users

First fetch the user data from rollbar:

node ./rollbar-stats.js <start date yyyy-mm-dd> [<end date yyyy-mm-dd>]

Examples:

node ./rollbar-stats.js 2014-07-27 - Gets data for the 2014-07-27 day.

node ./rollbar-stats.js 2014-07-25 2014-07-27 - Gets data for the 2014-07-25 - 2014-07-27 time period.

This command will create the users_<start date yyyy-mm-dd>_<end date yyyy-mm-dd>.json file and it will create or update the report-data.json file.

The format of the users_*.json file is:

{
    "username": [array of all rollbar instances timestamps],
    ...
}

The format of the report-data.json file is:

{
    "days": {
        "yyyy/mm/dd"
    }
    ...
}

Usernames in the ignore-usernames.txt file are ignored and will not be part of the json file created.

Then, execute the command that sill process the data and create the CSV files with the stats:

node ./build_report.js <users data json file>

Example:

node ./build_report.js users_2014-07-27.json

This will create or update the users.csv and activity.csv files.

Github

First fetch the user data from github:

node ./github-stats.js

This command will create or update the github-stats.json file.

The format of this file is:

{
    "users": {
        "username": {
            "avatar_url",
            "name",
            "company",
            "location",
            "email",
            "bio",
            "created_at",
            "public_repos",
            "repos": {
                "repo-name": {
                    "isTodoFork": <boolean>,
                    "isPopcornFork": <boolean>,
                    "masterNumberCommits": <number>,
                    "shadowNumberCommits": <number>
                },
                ...
            }
        },
        ...
    }
}

Then, execute the command that sill process the data and create the CSV files with the github stats:

node ./build-github-report.js github-stats.json

This command will create the github.csv file.

Content of the CSV files

users.csv

date - the date of the data, this is an entire day (12:00am to 11:59pm PST).

active_users - the number of unique users that used montage studio.

new_users - the number of users that used montage studio for the very first time.

activity.csv

This report uses the concept of active day. An active day is a day (a time period between 12:00am - 11:59pm PST) where the user opened montage studio at least once. It does not imply that the user actually spent time using montage studio.

username - the github username.

active_days - the total number of active days since account creation.

last_active_day - the last active day.

first_active_day - the first active day, this is the first time the user opened montage studio.

active_range_ratio - days_between(first_active_day, last_active_day) / days_between(first_active_day, today) - this is a tentative number for knowing when the person stopped using the website but by giving a bigger score to people who have used the site longer. If user A and B haven't used the website for the past 5 days but user A has used it for a total of 30 days while user B only used it for 10 days, user A will have a better score.

active_days_ratio - active_days / days_between(first_active_day, today) - the percentage of time the user has used the tool, since its first usage.

github.csv

user - the github username.

montage_repo - montage repository name.

new_commits - number of commits after master, trying to understand how many commits were made in montage studio.

new_github_user - if this user created a github account just to use montage studio. This is an heuristic, this is true when the github account creation matches the day of first usage of montage studio.